Like many folks who use three-tier design, I would like to create an audit trail in my Postgres database, and I would like to do so without having to create a database user for each audit.
As I see it, there are two ways to do this, and I can't see a clear way to do either of them. If anyone has better suggestions, I'd of course love to hear them.
Here's what I'd thought up:
1) Connect my homebrew login system which runs out of a couple database tables to postgres connection/sessionID (i.e., keep track of which sessionID represents my current user) so that any audit function can use the session ID to look up the current user.
2) Maintain a "current homebrew user" session variable that is distinct from Postgres' current_user, which I believe stores the current database user. I found a couple threads on session variables, but mostly they were discouraging people from using such variables.
Does anyone have any good ideas or advice?
Also, both of these methods require that a user maintain his/her own session. I don't know how PG's connection pooling works, but is it actually possible to specify a particular session for a particular user? Is there some place I can find documentation on how Postgres deals with logins and sessions?
Many thanks,
Eric
---------------------------(end of broadcast)---------------------------
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match
